DESCRIPTION:(adapted from applicant's abstract) In this proposal, it is hypothesized that during the development of habit learning, ensemble activity of striatal neurons undergoes systematic changes. It is proposed to test this hypothesis by using multiple tetrode chronic recording methods to record activity of striatal neurons during learning in rats, and by performing a set of analyses to characterize single-unit and ensemble activity in relation to task acquisition and performance. They propose to determine the factors responsible for changes in striatal activity during procedural learning in a t-maze, and to determine whether different striatal zones and corticostriatal loops undergo different patterns of plasticity during learning. They will also compare the response patterns in the striatum and hippocampus during acquisition of the cued "win-stay" versus memory based "win-shift" versions of the task. They will test the following hypotheses: (1) The large response to the go-signal that they have found to develop over learning is hypothesized to reflect a general readiness to respond rather than a response to detailed reward prediction or instructed behavior prediction. (2) The large changes that they have found to occur in turning responses are hypothesized to reflect the development of habit, and will not occur in the win-shift task. (3) They hypothesize that learning related changes will be selective for, and different in, the dorsolateral and medial caudoputamen and that these changes will be different from those in the corresponding cortical areas. (4) They hypothesize that during win-stay and win-shift t-maze learning, the dorsolateral caudoputamen and the hippocampus will code different classes of information, whereas the medial caudoputamen and the hippocampus will process similar information in different manners. They propose to test these hypotheses by using different versions of the t-maze task to vary predictability of reward or instructed behavior, to vary win-stay versus win-shift task requirements, by recording in different striatal sites and in pairs of striatal and cortical sites making up functionally distinct corticostriatal loops, and by comparing striatal and hippocampal neuronal activities during training on the win-stay and win-shift t-maze tasks.